diff --git a/app/chain/site.py b/app/chain/site.py index fd95e3915..b020589d8 100644 --- a/app/chain/site.py +++ b/app/chain/site.py @@ -203,6 +203,7 @@ class SiteChain(InteractionChainMixin, ChainBase): "hddolby.com": self.__hddolby_test, "rousi.pro": self.__rousi_test, "sunnypt.top": self.__sunnypt_test, + "milkie.cc": self.__milkie_test, } def refresh_userdata(self, site: dict = None) -> Optional[SiteUserData]: @@ -403,6 +404,33 @@ class SiteChain(InteractionChainMixin, ChainBase): return False, user_info.get("message", "鉴权已过期或无效") return False, f"错误：{res.status_code} {res.reason}" + def __milkie_test(self, site: SiteSnapshot) -> Tuple[bool, str]: + """ + 通过种子搜索接口测试 Milkie API Key + + :param site: Milkie 站点配置 + :return: 是否可用及状态信息 + """ + domain = site.domain or site_rules.extract_domain(site.url) + with _open_site_response( + method="GET", + url=f"https://{domain}/api/v1/torrents?ps=1", + headers={ + "Accept": "application/json", + "User-Agent": site.ua or self.runtime_config.user_agent, + "x-milkie-auth": site.apikey, + }, + proxies=(self.runtime_config.proxy or {}) if site.proxy else {}, + timeout=site.timeout or 15, + ) as res: + if res is None: + return False, "无法打开网站！" + if res.status_code == 200: + return True, "连接成功" + if res.status_code in (401, 403): + return False, "API Key 无效或已过期" + return False, f"错误：{res.status_code} {res.reason}" + def __sunnypt_test(self, site: SiteSnapshot) -> Tuple[bool, str]: """ 通过 profile 接口测试 SunnyPT API Key 和下载权限 diff --git a/app/modules/indexer/__init__.py b/app/modules/indexer/__init__.py index ef37d8312..cd9d227db 100644 --- a/app/modules/indexer/__init__.py +++ b/app/modules/indexer/__init__.py @@ -18,6 +18,7 @@ from app.modules.indexer.parser import SiteParserBase, SiteSchema from app.modules.indexer.spider import SiteSpider from app.modules.indexer.spider.haidan import HaiDanSpider from app.modules.indexer.spider.hddolby import HddolbySpider +from app.modules.indexer.spider.milkie import MilkieSpider from app.modules.indexer.spider.mtorrent import MTorrentSpider from app.modules.indexer.spider.rousi import RousiSpider from app.modules.indexer.spider.sunnypt import SunnyPTSpider @@ -39,6 +40,7 @@ SPIDER_PARSER_CLASSES = { "HDDolby": HddolbySpider, "RousiPro": RousiSpider, "SunnyPT": SunnyPTSpider, + "Milkie": MilkieSpider, } _SPECIALIZED_SEARCH_ARGUMENTS = { @@ -50,6 +52,7 @@ _SPECIALIZED_SEARCH_ARGUMENTS = { "Haidan": ("keyword", "mtype"), "HDDolby": ("keyword", "mtype", "page"), "RousiPro": ("keyword", "mtype", "cat", "page"), + "Milkie": ("keyword", "mtype", "page"), } _SPECIALIZED_SUBTITLE_SEARCH_ARGUMENTS = { diff --git a/app/modules/indexer/spider/milkie.py b/app/modules/indexer/spider/milkie.py new file mode 100644 index 000000000..2c100031a --- /dev/null +++ b/app/modules/indexer/spider/milkie.py @@ -0,0 +1,178 @@ +from typing import Any, List, Optional, Tuple +from urllib.parse import quote + +from app.adapters.network.http import AsyncRequestUtils, RequestUtils +from app.foundation import temporal as time_tools +from app.runtime.log import logger +from app.runtime.settings import get_runtime_setting +from app.schemas.types import MediaType + + +class MilkieSpider: + """Milkie JSON API 种子搜索适配器。""" + + _size = 100 + _timeout = 15 + _proxy: Optional[dict[str, Any]] = None + _searchurl = "%sapi/v1/torrents" + _downloadurl = "%sapi/v1/torrents/%s/torrent?key=%s" + _pageurl = "%sbrowse/%s" + + # 电影分类 + _movie_category = "1" + # 电视剧分类 + _tv_category = "2" + # 音乐分类 + _music_category = "3" + + # 站点分类到媒体类型的映射，未覆盖的分类不输出媒体类型 + _category_mapping = { + 1: MediaType.MOVIE, + 2: MediaType.TV, + 3: MediaType.MUSIC, + } + + def __init__(self, indexer: dict[str, Any]): + """使用站点配置初始化 Milkie 请求上下文。""" + if indexer: + self._indexerid = indexer.get('id') + self._domain = indexer.get('domain') + self._searchurl = self._searchurl % self._domain + self._name = indexer.get('name') + if indexer.get('proxy'): + self._proxy = get_runtime_setting('PROXY') + self._apikey = indexer.get('apikey') + self._ua = indexer.get('ua') + self._timeout = indexer.get('timeout') or 15 + + @classmethod + def get_search_page_size(cls, keyword: Optional[str] = None) -> Optional[int]: + """ + 获取搜索接口单页容量。 + """ + return cls._size + + def __get_params(self, keyword: Optional[str] = None, mtype: Optional[MediaType] = None, + page: Optional[int] = 0) -> dict[str, Any]: + """ + 获取搜索参数 + """ + params = { + "pi": int(page or 0), + "ps": self._size, + "oby": "created_at", + "odir": "desc" + } + if keyword: + params["query"] = keyword + # 按媒体类型收敛分类，提高关键字搜索的准确度 + if mtype == MediaType.MOVIE: + params["categories"] = self._movie_category + elif mtype == MediaType.TV: + params["categories"] = self._tv_category + elif mtype == MediaType.MUSIC: + params["categories"] = self._music_category + return params + + def __parse_created_at(self, created_at: str) -> str: + """将站点 ISO8601 发布时间转换为本地日期时间文本。""" + parsed = time_tools.parse_datetime(created_at) + if not parsed: + return created_at or "" + return parsed.strftime("%Y-%m-%d %H:%M:%S") + + def __parse_result(self, results: List[dict[str, Any]]) -> List[dict[str, Any]]: + """ + 解析搜索结果 + """ + torrents: List[dict[str, Any]] = [] + if not results: + return torrents + + for result in results: + category = self._category_mapping.get(result.get('category') or 0) + torrent = { + 'title': result.get('releaseName'), + 'description': result.get('id'), + # API Key 含 URL 保留字符，必须编码后拼入下载链接 + 'enclosure': self._downloadurl % ( + self._domain, result.get('id'), quote(self._apikey or "", safe="") + ), + 'pubdate': self.__parse_created_at(result.get('createdAt') or ""), + 'size': result.get('size'), + 'seeders': result.get('seeders'), + 'peers': result.get('leechers'), + 'grabs': result.get('downloaded'), + # 站点无分享率考核，等效全站免费 + 'downloadvolumefactor': 0, + 'uploadvolumefactor': 1, + 'page_url': self._pageurl % (self._domain, result.get('id')), + 'category': category.value if category else None + } + externals = result.get('externals') or {} + if externals.get('imdb'): + torrent['imdbid'] = externals.get('imdb') + torrents.append(torrent) + + return torrents + + def __process_response(self, res: Any) -> Tuple[bool, List[dict[str, Any]]]: + """统一判定搜索响应状态并投影 Milkie 种子结果。""" + if res and res.status_code == 200: + results = res.json().get("torrents") or [] + return False, self.__parse_result(results) + if res is not None: + if res.status_code in (401, 403): + logger.warn(f"{self._name} 搜索失败，API Key 无效或已过期") + else: + logger.warn(f"{self._name} 搜索失败，错误码：{res.status_code}") + return True, [] + logger.warn(f"{self._name} 搜索失败，无法连接 {self._domain}") + return True, [] + + def __get_headers(self) -> dict[str, str]: + """构造带 API Key 认证的请求头。""" + return { + "x-milkie-auth": self._apikey or "", + "User-Agent": self._ua or "" + } + + def search(self, keyword: str, mtype: Optional[MediaType] = None, + page: Optional[int] = 0) -> Tuple[bool, List[dict[str, Any]]]: + """ + 搜索 + """ + if not self._apikey: + logger.warn(f"{self._name} 未配置 API Key，无法搜索") + return True, [] + + # 获取请求参数 + params = self.__get_params(keyword, mtype, page) + + # 发送请求 + res = RequestUtils( + headers=self.__get_headers(), + proxies=self._proxy, + timeout=self._timeout + ).get_res(url=self._searchurl, params=params) + return self.__process_response(res) + + async def async_search(self, keyword: str, mtype: Optional[MediaType] = None, + page: Optional[int] = 0) -> Tuple[bool, List[dict[str, Any]]]: + """ + 异步搜索 + """ + if not self._apikey: + logger.warn(f"{self._name} 未配置 API Key，无法搜索") + return True, [] + + # 获取请求参数 + params = self.__get_params(keyword, mtype, page) + + # 发送请求 + res = await AsyncRequestUtils( + headers=self.__get_headers(), + proxies=self._proxy, + timeout=self._timeout + ).get_res(url=self._searchurl, params=params) + return self.__process_response(res) diff --git a/tests/fixtures/architecture/dependency-baseline.json b/tests/fixtures/architecture/dependency-baseline.json index 5840a37a3..9d9d11e67 100644 --- a/tests/fixtures/architecture/dependency-baseline.json +++ b/tests/fixtures/architecture/dependency-baseline.json @@ -941,8 +941,8 @@ "runtime_only": true } }, - "edge_count": 9401, - "edge_sha256": "82e88feed7d0c88ac1617017a301ed2979be7da87e995463b5e22b61c77e9620", + "edge_count": 9412, + "edge_sha256": "12ffbe3a91e7314ded4ce7996bffeb7497ba0e2434bb57e58e428da20dfd0ed6", "edges": [ "app -> app.foundation", "app -> app.foundation.environment", @@ -7405,6 +7405,7 @@ "app.modules.indexer -> app.modules.indexer.spider", "app.modules.indexer -> app.modules.indexer.spider.haidan", "app.modules.indexer -> app.modules.indexer.spider.hddolby", + "app.modules.indexer -> app.modules.indexer.spider.milkie", "app.modules.indexer -> app.modules.indexer.spider.mtorrent", "app.modules.indexer -> app.modules.indexer.spider.rousi", "app.modules.indexer -> app.modules.indexer.spider.sunnypt", @@ -7628,6 +7629,16 @@ "app.modules.indexer.spider.hddolby -> app.runtime.settings", "app.modules.indexer.spider.hddolby -> app.schemas", "app.modules.indexer.spider.hddolby -> app.schemas.types", + "app.modules.indexer.spider.milkie -> app.adapters", + "app.modules.indexer.spider.milkie -> app.adapters.network", + "app.modules.indexer.spider.milkie -> app.adapters.network.http", + "app.modules.indexer.spider.milkie -> app.foundation", + "app.modules.indexer.spider.milkie -> app.foundation.temporal", + "app.modules.indexer.spider.milkie -> app.runtime", + "app.modules.indexer.spider.milkie -> app.runtime.log", + "app.modules.indexer.spider.milkie -> app.runtime.settings", + "app.modules.indexer.spider.milkie -> app.schemas", + "app.modules.indexer.spider.milkie -> app.schemas.types", "app.modules.indexer.spider.mtorrent -> app.adapters", "app.modules.indexer.spider.mtorrent -> app.adapters.network", "app.modules.indexer.spider.mtorrent -> app.adapters.network.http", @@ -10346,7 +10357,7 @@ "app.workflow.actions.transfer_file -> app.workflow", "app.workflow.actions.transfer_file -> app.workflow.actions" ], - "module_count": 1112, + "module_count": 1113, "modules": [ "app", "app.adapters", @@ -11121,6 +11132,7 @@ "app.modules.indexer.spider", "app.modules.indexer.spider.haidan", "app.modules.indexer.spider.hddolby", + "app.modules.indexer.spider.milkie", "app.modules.indexer.spider.mtorrent", "app.modules.indexer.spider.rousi", "app.modules.indexer.spider.sunnypt", diff --git a/tests/test_indexer_sync_async_parity.py b/tests/test_indexer_sync_async_parity.py index 8859bfdfe..27b48e488 100644 --- a/tests/test_indexer_sync_async_parity.py +++ b/tests/test_indexer_sync_async_parity.py @@ -89,6 +89,10 @@ def _install_search_observers(monkeypatch, events: list[tuple]) -> None: "Yema", {"keyword": "clean keyword", "mtype": MediaType.MOVIE, "page": 3}, ), + ( + "Milkie", + {"keyword": "clean keyword", "mtype": MediaType.MOVIE, "page": 3}, + ), ( "Haidan", {"keyword": "clean keyword", "mtype": MediaType.MOVIE}, diff --git a/tests/test_milkie_spider.py b/tests/test_milkie_spider.py new file mode 100644 index 000000000..465f43e43 --- /dev/null +++ b/tests/test_milkie_spider.py @@ -0,0 +1,205 @@ +# -*- coding: utf-8 -*- +# pylint: disable=no-name-in-module + +from unittest.mock import AsyncMock, MagicMock + +import pytest + +from app.modules.indexer.spider import milkie as milkie_module +from app.modules.indexer.spider.milkie import MilkieSpider +from app.schemas.types import MediaType + + +def _build_indexer() -> dict: + """构造 Milkie API Spider 所需的最小站点配置。""" + return { + "id": "milkie", + "name": "Milkie", + "domain": "https://milkie.cc/", + "apikey": "milkie+secret", + "ua": "MoviePilot-Test", + "proxy": False, + } + + +def _sample_torrent() -> dict: + """构造 Milkie 搜索接口的一条真实形态结果。""" + return { + "id": "wrJoUrlySnVm", + "releaseName": "Dune.Part.Two.2024.2160p.UHD.BluRay.H265-GAZPROM", + "category": 1, + "createdAt": "2026-10-08T08:46:25+02:00", + "size": 1076476044, + "downloaded": 12, + "seeders": 11, + "leechers": 0, + "externals": {"imdb": "tt15239678", "tmdb": None, "tvmaze": None}, + } + + +@pytest.fixture() +def milkie_spider(): + """构造不依赖真实站点配置的 MilkieSpider。""" + return MilkieSpider(_build_indexer()) + + +def test_search_page_size_matches_api_limit(milkie_spider): + """单页容量应与接口上限一致，供分页器正确切片。""" + assert MilkieSpider.get_search_page_size("dune") == 100 + + +def test_keyword_search_maps_media_categories(milkie_spider): + """带关键字的搜索应按媒体类型收敛分类。""" + params = milkie_spider._MilkieSpider__get_params( + "Dune", MediaType.MOVIE, 0 + ) + + assert params["query"] == "Dune" + assert params["categories"] == "1" + assert params["pi"] == 0 + assert params["ps"] == 100 + + +def test_tv_and_music_category_mapping(milkie_spider): + """剧集与音乐分别映射到站点 2、3 分类。""" + tv_params = milkie_spider._MilkieSpider__get_params("Dune", MediaType.TV, 1) + music_params = milkie_spider._MilkieSpider__get_params("Dune", MediaType.MUSIC, 0) + + assert tv_params["categories"] == "2" + assert tv_params["pi"] == 1 + assert music_params["categories"] == "3" + + +def test_keyword_search_without_media_type_omits_categories(milkie_spider): + """未指定媒体类型时不提交分类，避免站点接口过滤失效。""" + params = milkie_spider._MilkieSpider__get_params("Dune", None, 0) + + assert "categories" not in params + + +def test_browse_mode_supports_pagination(milkie_spider): + """无关键字的浏览模式同样支持服务端翻页。""" + first_page = milkie_spider._MilkieSpider__get_params(None, None, 0) + second_page = milkie_spider._MilkieSpider__get_params(None, None, 1) + + assert "query" not in first_page + assert first_page["pi"] == 0 + assert second_page["pi"] == 1 + assert first_page["oby"] == "created_at" + + +def test_parse_result_projects_torrent_fields(milkie_spider): + """结果字段应完整映射到站点搜索契约。""" + torrents = milkie_spider._MilkieSpider__parse_result([_sample_torrent()]) + + assert len(torrents) == 1 + torrent = torrents[0] + assert torrent["title"] == "Dune.Part.Two.2024.2160p.UHD.BluRay.H265-GAZPROM" + assert torrent["enclosure"] == ( + "https://milkie.cc/api/v1/torrents/wrJoUrlySnVm" + "/torrent?key=milkie%2Bsecret" + ) + assert torrent["page_url"] == "https://milkie.cc/browse/wrJoUrlySnVm" + assert torrent["size"] == 1076476044 + assert torrent["seeders"] == 11 + assert torrent["peers"] == 0 + assert torrent["grabs"] == 12 + assert torrent["pubdate"] == "2026-10-08 08:46:25" + assert torrent["imdbid"] == "tt15239678" + assert torrent["category"] == MediaType.MOVIE.value + + +def test_parse_result_marks_site_as_freeleech(milkie_spider): + """站点无分享率考核，所有种子应视为免费。""" + torrents = milkie_spider._MilkieSpider__parse_result([_sample_torrent()]) + + assert torrents[0]["downloadvolumefactor"] == 0 + assert torrents[0]["uploadvolumefactor"] == 1 + + +def test_parse_result_category_fallback(milkie_spider): + """游戏、软件等无对应媒体类型的分类不应输出媒体类型。""" + raw = _sample_torrent() + raw["category"] = 4 + + torrents = milkie_spider._MilkieSpider__parse_result([raw]) + + assert torrents[0]["category"] is None + + +def test_search_requires_api_key(milkie_spider, monkeypatch): + """缺少 API Key 时应直接失败且不发起请求。""" + request_utils = MagicMock() + monkeypatch.setattr(milkie_module, "RequestUtils", request_utils) + + milkie_spider._apikey = None + error, results = milkie_spider.search("Dune") + + assert error is True + assert results == [] + request_utils.assert_not_called() + + +def _install_json_response(monkeypatch, status_code: int, payload: dict) -> MagicMock: + """在模块边界替换同步 HTTP 客户端并返回调用记录。""" + response = MagicMock() + response.status_code = status_code + response.json.return_value = payload + request_utils = MagicMock() + request_utils.return_value.get_res.return_value = response + monkeypatch.setattr(milkie_module, "RequestUtils", request_utils) + return request_utils + + +def test_search_sends_auth_header_and_parses_response(milkie_spider, monkeypatch): + """搜索应携带 x-milkie-auth 请求头并解析 torrents 数组。""" + request_utils = _install_json_response( + monkeypatch, 200, {"hits": 1, "torrents": [_sample_torrent()]} + ) + + error, results = milkie_spider.search("Dune", MediaType.MOVIE) + + assert error is False + assert len(results) == 1 + headers = request_utils.call_args.kwargs["headers"] + assert headers["x-milkie-auth"] == "milkie+secret" + + +def test_search_reports_invalid_api_key(milkie_spider, monkeypatch): + """401 响应应判定为 API Key 失效。""" + _install_json_response(monkeypatch, 401, {"message": "Unauthorized"}) + + error, results = milkie_spider.search("Dune") + + assert error is True + assert results == [] + + +def test_search_reports_unreachable_site(milkie_spider, monkeypatch): + """网络不可达时应返回错误标志。""" + request_utils = MagicMock() + request_utils.return_value.get_res.return_value = None + monkeypatch.setattr(milkie_module, "RequestUtils", request_utils) + + error, results = milkie_spider.search("Dune") + + assert error is True + assert results == [] + + +@pytest.mark.asyncio +async def test_async_search_matches_sync_behavior(milkie_spider, monkeypatch): + """异步搜索应与同步搜索使用同样的认证与解析路径。""" + response = MagicMock() + response.status_code = 200 + response.json.return_value = {"hits": 1, "torrents": [_sample_torrent()]} + async_request_utils = MagicMock() + async_request_utils.return_value.get_res = AsyncMock(return_value=response) + monkeypatch.setattr(milkie_module, "AsyncRequestUtils", async_request_utils) + + error, results = await milkie_spider.async_search("Dune", MediaType.MOVIE) + + assert error is False + assert results[0]["title"] == "Dune.Part.Two.2024.2160p.UHD.BluRay.H265-GAZPROM" + headers = async_request_utils.call_args.kwargs["headers"] + assert headers["x-milkie-auth"] == "milkie+secret"